Real-time anti-fraud method and system based on network data and storage medium

Abstract

The invention provides a real-time anti-fraud method and system based on network data, which collects original network data and obtains user transaction information and user behavior information through receiving and analyzing the original network data. Sending user transaction information and user behavior information to fraud model matching; User transaction information and user behavior information are analyzed and matched by rule matching engine combined with machine learning fraud model, and the fraud behavior is judged according to the output processing result of fraud model. Judging thefraudulent behavior of the output of the fraudulent model to form the concrete fraudulent behavior; The data supply interface is invoked to send the concrete fraud behavior to the external processingterminal. Through the real-time collection and analysis of the network data, the invention directly obtains the network layer information without modifying the user service system, analyzes the network data with high performance, and outputs the fraudulent behavior result to the anti-fraudulent platform. Real-time, full-scale fraud detection data and analysis results are provided.

Description

technical field [0001] The invention relates to the field of real-time anti-fraud, in particular to a network data-based real-time fraud sending method, system and storage medium. Background technique [0002] With the Internet-based business of traditional financial institutions and the rapid development of new Internet financial models, Internet fraud methods have become more and more diverse, and the risk of fraud for users and Internet financial platforms has also increased sharply. Statistics show that China's Internet fraud risk ranks among the top three in the world, and the loss caused by Internet fraud has reached 0.63% of GDP, which is second only to the 0.64% of the United States. [0003] The current common anti-fraud data acquisition methods obtain information through databases or logs or embedded mode, which is not effective or comprehensive enough. By obtaining data in the database from the database, the timeliness is poor, and the transaction process data wil...

AI Technical Summary

Problems solved by technology

The above patent literature uses the obtained data sources to perform anti-fraud calculations, but only uses historical data sources for calculations, which is not comprehensive enough

Images